Vegetation Reconstruction Method In Lingquan Open-pit Mine Based On Vegetation Classification And Site Type Division

Coal resources play a very important role and have made significant contributions to the recovery and rapid development of the national economy.The continuous mining of open-pit coal mines has caused varying degrees of damage to the original vegetation in the mining area and its surrounding areas,leading to retrogressive succession of vegetation and ecological degradation to some extent.The cold climate,large diurnal temperature range,and short growing season of plants in high-altitude grasslands make ecological restoration a slow process.Therefore,the reconstruction of suitable vegetation in the area is helpful in accelerating the ecological restoration of highaltitude open-pit coal mines.However,there is currently limited research on vegetation reconstruction in high-altitude grasslands,and there is a lack of relevant restoration cases and experience,making the research on vegetation reconstruction a current priority.Since the complete closure of Lingquan open-pit mine in 2017,some areas have yet to undergo vegetation reconstruction work.Optimizing the vegetation reconstruction method based on previous restoration experience is of great significance to accelerate the restoration process in the mining area.This study takes Lingquan openpit mine,the first closed open-pit coal mine in China’s high-altitude grassland region,as an example to conduct research on vegetation reconstruction in high-altitude openpit coal mines.The main research contents and results are as follows:(1)Obtaining prior knowledge on vegetation classification based on highresolution remote sensing imagesThe object-oriented support vector machine(SVM)method was used to classify vegetation in the mining area and compared with pixel-based SVM method,pixel-based random forest classification,and pixel-based maximum likelihood classification accuracy.The results showed that the object-oriented SVM classification method was suitable for the vegetation classification of the Lingquan open-pit mine.Combining with the actual situation of the mining area,the vegetation in the mining area from 2017 to 2022 was classified into trees,shrubs,and herbs using this classification method.The results showed that the area cover by vegetation was increasing year by year,and the obtained vegetation classification results can be used as prior knowledge for vegetation reconstruction in the mining area.(2)Prior knowledge acquisition of site typesCombining in-situ sample measurements with slope,aspect,and other knowledge obtained from remote sensing images,slope,gravel content,soil water content,and soil organic matter content were considered as site factors for dividing the mining area into different site types.Correlation analysis and principal component analysis were used to determine the dominant factors in dividing site types.Based on this,and considering the actual situation of the mining area,the mining area was divided into seven site types:slightly sloping low fertility site,slightly sloping medium fertility site,slightly sloping high fertility site,gentle sloping low fertility site,gentle sloping medium fertility site,steep sloping medium fertility site,and steep sloping high fertility site.The corresponding site characteristics were analyzed,and the obtained site types can be used as prior knowledge for mining area vegetation reconstruction.(3)Vegetation reconstruction of Lingquan open-pit mining area based on decision tree model with prior knowledgeTaking different site types as units,the growth of corresponding vegetation types was analyzed.The results show that the effect of vegetation reconstruction in mining area is remarkable in recent years.Therefore,the experience of vegetation reconstruction in mining area in recent years can be used as reference for vegetation reconstruction in unrepaired areas of mining area.Samples were selected from areas with good vegetation growth,and the vegetation classification results(trees,shrubs,and herbs)can be used as decision attributes,while site types were used as conditional attributes.Combining the species diversity and dominant species investigated in different site types,a vegetation restoration method based on a decision tree model was obtained.This method can be used as a reference for the vegetation restoration of unrepaired areas in the high-altitude grassland open-pit coal mines under similar site conditions.